Two million Fires annually kill and injure more than 27,000 people in the US. All other natural disasters (floods, hurricanes, earthquakes, tornadoes, etc) account for only a small percentage of those injured in fire every year and it is estimated that as many as 200,000-300,000 are actually injured in fire but the majority are not reported. Fire is the 5th leading cause of accidental death behind only motor vehicle accidents, falls, poisonings, and drowning deaths. In addition fire causes over $9 billion in direct damage and in excess of $100 billion in total damages. A significant decrease in both injuries and the monetary loss from fire has occurred over the past twenty years and this is credited in large part to the widespread use of smoke detectors. Smoke detectors and smoke alarm systems not only provide a means of reducing death and injury, but also allow for a quicker response to the fire, resulting in less damage. The goal of this research project is to develop a portable, inexpensive device for use by the hearing impaired community (approximate 22 million in the United States [Adams]) that identifies the audio signal from a smoke detector, a fire alarm horn or other emergency notification audio signal and then activates an alert mechanism capable of waking a deaf individual in all stages of sleep. The aim is for the devices signal recognition capabilities of the device to provide equal protection to this disabled population as that provided to the hearing able population as dictated in the National Fire Protection Association 72, National Fire Alarm Code. These concepts will be investigated in two distinct paths. The first path will further refine the auditory recognition capabilities of a signal identification algorithm developed in Phase I research to include nonperiodic smoke detectors and emergency alarm horns. The second path of this research involves testing deaf, hearing-impaired and hearing able individuals for their response to alert mechanisms. The objective is to determine the effectiveness of audible smoke detectors in waking hearing able individuals and use this information as a base line for deaf individuals response to tactile or visual waking devices. No research could be identified in the literature that links sleep stage monitoring and effectiveness of waking of emergency alarms. The goal of Phase II is to arrive at a suitable "brass board" prototype device. [unreadable] [unreadable]